In order to solve the problem of poor performance of unmanned rescue flight control system,the flight control system of unmanned rescue has been designed for lift of its flight stability.A flight control algorithm suitable for flight control system of the intelligent search and rescue UAV has also been proposed based on LM-Smith neural network,which mainly includes two parts,design of control algorithm and that of hardware module.The LM-Smith neural network functional weighted method with minimum information entropy has been employed for pattern recognition of controlled object and self-adaptive modification of flight stability.The location offsets in process of flight control has been tracked and compensated wholly.And thus asymptotically stable convergence of the proposed flight control algorithm has been achieved via established functional controller with LM-Smith neural network.The promoted design of flight control system has been implemented by adopting S3C2440A processor as control core and processing and ARM9 as core main contro1 board and loading embedded Linux system as control program.The tests and analysis have been conducted to verify effectiveness and applicability of both the established flight control system and the designed flight control algorithm.The test results show that the established flight control system has higher precision of trajectory tracking and control,better attitude stability and control performance.%针对当前搜救无人机飞控系统控制性能差的问题,基于嵌入式平台进行了搜救无人机的飞控系统设计,以提高搜救无人机的飞行稳定性,并提出一种基于LM-Smith神经网络泛函的搜救无人机飞行控制算法.飞控系统设计主要包括控制算法设计和硬件模块设计两大部分,采用最小信息熵进行LM-Smith神经网络泛函加权,对被控对象进行模式识别和稳定性自适应修正,对飞行控制的位置偏移进行完全跟踪补偿,通过构建的LM-Smith神经网络泛函的控制器,获得渐进稳定收敛的飞行控制算法.以S3C2440A ARM9处理器作为控制核心和以Micro2440为核心主控板,基于嵌入式Linux系统加载控制程序,实现了控制系统改进设计.为验证飞控系统与控制算法的有效性和适用性,进行了性能测试与分析.测试结果表明,所构建的飞控系统在智能搜救无人机飞行控制中具有较高的轨迹跟踪控制精度,姿态稳定性和控制性能较好.
展开▼
